half SBS vs full - I know it's less but can you tell the difference? - AVS Forum
Forum Jump: 
Reply
 
Thread Tools
post #1 of 74 Old 11-19-2010, 04:45 AM - Thread Starter
Advanced Member
 
delt31's Avatar
 
Join Date: Jun 2007
Posts: 628
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
I was wondering if you can tell the difference when watching half SBS vs full 3d? Anyone with experience watching half sbs and a full 3d bluray on ps3 or something similar?
delt31 is offline  
Sponsored Links
Advertisement
 
post #2 of 74 Old 11-19-2010, 06:17 AM
Member
 
joels1010's Avatar
 
Join Date: Mar 2001
Posts: 184
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
I really cant tell the difference, but i know people will chime in to a different tune.
joels1010 is offline  
post #3 of 74 Old 11-19-2010, 08:24 AM
AVS Special Member
 
Frank's Avatar
 
Join Date: Apr 1999
Location: Dept. of Offense
Posts: 5,194
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 4 Post(s)
Liked: 14
It depends.
Since full 1080P per eye only works at 24 frames per second,
a 24fps gaussian blur must be applied to motion.
If you find a single frame with little or no motion and view it both ways then you can "probably" tell the difference.

Did IQ's suddenly drop sharply while I was away?
I enjoy 3D in spite of HDMI 1.4!
Full screen only 3D doesn't cut it!
Frank is offline  
post #4 of 74 Old 11-19-2010, 09:38 AM
 
icester's Avatar
 
Join Date: Apr 2010
Posts: 1,175
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by delt31 View Post

I was wondering if you can tell the difference when watching half SBS vs full 3d? Anyone with experience watching half sbs and a full 3d bluray on ps3 or something similar?

There is no difference if both files have the same size.
Also the only real full HD 3D is generated by Games using nVidia's GPU.
The 3D-Blu-Ray's 2D + Delta format is just a gimmick.
Frame compatible Side By Side will deliver the same quality or even better if compression ratios are the same or the file size are the same.

Mathew Orman

http://www.*******************.com/
icester is offline  
post #5 of 74 Old 11-19-2010, 12:40 PM
AVS Addicted Member
 
walford's Avatar
 
Join Date: May 2003
Location: Orange County, CA
Posts: 16,789
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
Why arn't 3D BR MVD encoded disks that pass the MVC decoding tests supply true 3D when decoded and output using the double buffer 1080p per eye HDMI 1.4a 3d format equal or better then instead of 720p per eye frame sequential/page flipping provided by the Nvidia 3D player?
walford is offline  
post #6 of 74 Old 11-19-2010, 12:52 PM
 
icester's Avatar
 
Join Date: Apr 2010
Posts: 1,175
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by walford View Post

Why arn't 3D BR MVD encoded disks that pass the MVC decoding tests supply true 3D when decoded and output using the double buffer 1080p per eye HDMI 1.4a 3d format instead of 720p per eye frame sequential/page flipping provided by the Nvidia 3D player?

What kind of English is that?

Mathew Orman
icester is offline  
post #7 of 74 Old 11-19-2010, 01:41 PM
 
Lee Stewart's Avatar
 
Join Date: Feb 2007
Location: Albuquerque, NM
Posts: 19,369
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 44
Quote:
Originally Posted by icester View Post

The 3D-Blu-Ray's 2D + Delta format is just a gimmick.

Mathew Orman

http://www.*******************.com/

Please explain that statement.
Lee Stewart is offline  
post #8 of 74 Old 11-19-2010, 06:01 PM
Advanced Member
 
BlackShark's Avatar
 
Join Date: Jun 2010
Location: Montpellier, France
Posts: 580
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 13
Quote:
Originally Posted by icester View Post

There is no difference if both files have the same size.
Also the only real full HD 3D is generated by Games using nVidia's GPU.
The 3D-Blu-Ray's 2D + Delta format is just a gimmick.
Frame compatible Side By Side will deliver the same quality or even better if compression ratios are the same or the file size are the same.

Mathew Orman

http://www.*******************.com/

I disagree with EVERY sentence you wrote here.

I have 3 BluRay3D movies, they're all stereo FullHD (or the cinemascope equivalent) I also have an ATI GPU and I play my games in FullHD 3D on my dual projectors

MPEG4 MVC's 2D+Delta is the real deal and is the most efficient stereo video compresison format to date, if you have a good encoder you will get better quality for the file size than anything else for stereo video.

There is a difference between full and half res side by side, and yes of course it is visible, but the obviousness depends on the quality of the down and upscaling performed as well as the video compression format.

Compression ratios can be used to compare image quality between a source and it's compressed file, but not between compressed files of different sources.
Different sources compressed with the same compression ratios will also have different quality thus cannot be used directly to compare
If both full res and half res files have the same file size, they have different picture quality

With modern encoders, the picture quality and compression ratios are not proportional, so unless you are using extremely low bitrates if the two files have the same file size, the full res picture will look better visually than the half res picture.

Passive 3D, forever !
My Full-HD dual-projector passive polarised 3D setup. (really out of date ! I need to update it some day...)

BlackShark is offline  
post #9 of 74 Old 11-19-2010, 06:38 PM
AVS Addicted Member
 
walford's Avatar
 
Join Date: May 2003
Location: Orange County, CA
Posts: 16,789
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
Quote:
Originally Posted by icester View Post

What kind of English is that?

Mathew Orman

US Engish by senior citizens not the Kings English as used in England.
But my poor keyboard senior citizen US english is no excuse for your totally ignoring the question I asked.
I have edited by post in hopes that you can understand it better.
walford is offline  
post #10 of 74 Old 11-20-2010, 03:37 AM
 
icester's Avatar
 
Join Date: Apr 2010
Posts: 1,175
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by walford View Post

US Engish by senior citizens not the Kings English as used in England.
But my poor keyboard senior citizen US english is no excuse for your totally ignoring the question I asked.
I have edited by post in hopes that you can understand it better.

My apologies.
Can you break down the question to less complex one.
I have a hard time to provide an answer to such question.

Mathew Orman
icester is offline  
post #11 of 74 Old 11-20-2010, 03:50 AM
 
icester's Avatar
 
Join Date: Apr 2010
Posts: 1,175
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by BlackShark View Post

I disagree with EVERY sentence you wrote here.

I have 3 BluRay3D movies, they're all stereo FullHD (or the cinemascope equivalent) I also have an ATI GPU and I play my games in FullHD 3D on my dual projectors

MPEG4 MVC's 2D+Delta is the real deal and is the most efficient stereo video compresison format to date, if you have a good encoder you will get better quality for the file size than anything else for stereo video.

There is a difference between full and half res side by side, and yes of course it is visible, but the obviousness depends on the quality of the down and upscaling performed as well as the video compression format.

Compression ratios can be used to compare image quality between a source and it's compressed file, but not between compressed files of different sources.
Different sources compressed with the same compression ratios will also have different quality thus cannot be used directly to compare
If both full res and half res files have the same file size, they have different picture quality

With modern encoders, the picture quality and compression ratios are not proportional, so unless you are using extremely low bitrates if the two files have the same file size, the full res picture will look better visually than the half res picture.

It is just your opinion based on what you have to play with.
The real truth comes out when one actually does a practical prove of concept using a raw stereoscopic footage as an input.

Mathew Orman
icester is offline  
post #12 of 74 Old 11-20-2010, 04:55 AM
AVS Addicted Member
 
walford's Avatar
 
Join Date: May 2003
Location: Orange County, CA
Posts: 16,789
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
Quote:
Originally Posted by icester View Post

My apologies.
Can you break down the question to less complex one.
I have a hard time to provide an answer to such question.

Mathew Orman

Please explin why 3D games played using Nvidia 3D vision are true 3D and movies filmed with IMAX 3D cameras and show at an IMAX 3D theater maybe but if they are enocoded using (2D+Delta) format on a blu-Ray disk they are not?
walford is offline  
post #13 of 74 Old 11-20-2010, 05:01 AM
AVS Addicted Member
 
walford's Avatar
 
Join Date: May 2003
Location: Orange County, CA
Posts: 16,789
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
Quote:
Originally Posted by icester View Post

My apologies.
Can you break down the question to less complex one.
I have a hard time to provide an answer to such question.

Mathew Orman

Please explain why 3D games played using Nvidia 3D vision are "true 3D" and movies filmed with IMAX 3D cameras and shown at an IMAX 3D theater maybe but if they are "true 3D and then enocded using MVC(2D+Delta) format on a blu-Ray disk they are not any more?
walford is offline  
post #14 of 74 Old 11-20-2010, 05:59 AM
 
icester's Avatar
 
Join Date: Apr 2010
Posts: 1,175
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by walford View Post

Please explain why 3D games played using Nvidia 3D vision are "true 3D" and movies filmed with IMAX 3D cameras and shown at an IMAX 3D theater maybe but if they are "true 3D and then enocded using MVC(2D+Delta) format on a blu-Ray disk they are not any more?

Simply, nVidia outputs 1920 x 1080 x 32bit color, 60 Hz
without any compression in-between the card and monitor.
Encoded movies have reduced resolution due to lossy compression, motion blurring and camera focus blurring.
Additionally 3D movies encoded using MVC 2D + Delta method have reduced texture specularities and reduced or totally removed low contrast details.

Mathew Orman
icester is offline  
post #15 of 74 Old 11-20-2010, 12:05 PM
Advanced Member
 
BlackShark's Avatar
 
Join Date: Jun 2010
Location: Montpellier, France
Posts: 580
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 13
LoL icester,
Video games are not perfect sources, they have lots of shader approximations all over, blurry textures, reduced resolution (shadows and reflections), insufficient geometry details (obvious in close ups), insufficient rendering resolution (visible aliasing even with the best anti-aliasing modes enabled) in order to run the games in real time and most recent games even include motion blur and depth of field effects.

If the picture quality of 3D BluRays with MPEG4 MVC is insufficient for you then may I dare ask your opinion on 2D movies ?
2D BluRay (Mpeg4 AVC) ?
Digital cinema (jpeg2000) ?
Film ?

Passive 3D, forever !
My Full-HD dual-projector passive polarised 3D setup. (really out of date ! I need to update it some day...)

BlackShark is offline  
post #16 of 74 Old 11-20-2010, 01:05 PM
Member
 
Join Date: May 2010
Posts: 173
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
by 2014 every graphically, or just computationally intensive task will be run "on cloud" , this means limitless, supercomputer perfomance. Real time ray tracing will be commonplace very soon, streaming thru web browsers.

The console driven "prevgenHD" era is about to end. Now lets hope for MEMS displays.
gain3 is offline  
post #17 of 74 Old 11-20-2010, 03:44 PM
AVS Addicted Member
 
walford's Avatar
 
Join Date: May 2003
Location: Orange County, CA
Posts: 16,789
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
Quote:
Originally Posted by icester View Post

Simply, nVidia outputs 1920 x 1080 x 32bit color, 60 Hz
without any compression in-between the card and monitor.
Encoded movies have reduced resolution due to lossy compression, motion blurring and camera focus blurring.
Additionally 3D movies encoded using MVC 2D + Delta method have reduced texture specularities and reduced or totally removed low contrast details.

Mathew Orman

Do you consider programs "filmed" with 3D IMAX cameras and shown at 3D IMAX theaters as "true 3D" or as just a "gimmick"?

Doesn't your statement that "MVC 2D + Delta method have reduced texture specularities and reduced or totally removed low contrast details" depend on the quality of the MVC encoder or is true of all MVC encoders as you imply?
walford is offline  
post #18 of 74 Old 11-20-2010, 04:39 PM
 
icester's Avatar
 
Join Date: Apr 2010
Posts: 1,175
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by walford View Post

Do you consider programs "filmed" with 3D IMAX cameras and shown at 3D IMAX theaters as "true 3D" or as just a "gimmick"?

Doesn't your statement that "MVC 2D + Delta method have reduced texture specularities and reduced or totally removed low contrast details" depend on the quality of the MVC encoder or is true of all MVC encoders as you imply?

Since it's inception in 1971 IMAX totally relied on gimmick
that is an adrenaline WOW effect on viewers due to gigantism of scenes and actors in every movie.
It goes for both in 2D and 3D movies. Additionally IMAX productions of 3D movies are no different than competition and it also uses parallax limit driven stereoscopic content distortion techniques which yield totally unrealistic immersion.

Again MVC 2D + Delta technique was originally created in Germany's Fraunhofer Institute.

http://www.hhi.fraunhofer.de/en/depa...sing/overview/

The intention was to create efficient 3D Television signal compression method. It was established that in order to satisfy limited bandwidth compromises such as stereoscopic detail reduction had to be implemented.
The creators also acknowledge that the methods creates 3D artifacts that are not present in original content.

For those who would like to test the MVC codec and it's 3D implementation, here is the source code:

http://iphome.hhi.de/suehring/tml/download/

Mathew Orman
icester is offline  
post #19 of 74 Old 11-20-2010, 04:58 PM
 
icester's Avatar
 
Join Date: Apr 2010
Posts: 1,175
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by BlackShark View Post

LoL icester,
Video games are not perfect sources, they have lots of shader approximations all over, blurry textures, reduced resolution (shadows and reflections), insufficient geometry details (obvious in close ups), insufficient rendering resolution (visible aliasing even with the best anti-aliasing modes enabled) in order to run the games in real time and most recent games even include motion blur and depth of field effects.

If the picture quality of 3D BluRays with MPEG4 MVC is insufficient for you then may I dare ask your opinion on 2D movies ?
2D BluRay (Mpeg4 AVC) ?
Digital cinema (jpeg2000) ?
Film ?

That is nothing to do with Nvidia's GPU performance and it is only dependent on Game design.
4k textures can be used in rendering of FULL 3D HD format
without any compression between the card and display monitor.

Blue Ray has no magic capabilities and it is not capable of streaming real full HD content in 2D and much less in 3D.

Mathew Orman
icester is offline  
post #20 of 74 Old 11-20-2010, 05:03 PM
 
icester's Avatar
 
Join Date: Apr 2010
Posts: 1,175
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by gain3 View Post

by 2014 every graphically, or just computationally intensive task will be run "on cloud" , this means limitless, supercomputer perfomance. Real time ray tracing will be commonplace very soon, streaming thru web browsers.

The console driven "prevgenHD" era is about to end. Now lets hope for MEMS displays.

You are very optimistic.

Mathew Orman
icester is offline  
post #21 of 74 Old 11-20-2010, 05:15 PM
Advanced Member
 
BlackShark's Avatar
 
Join Date: Jun 2010
Location: Montpellier, France
Posts: 580
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 13
Quote:
Originally Posted by icester View Post

Again MVC 2D + Delta technique was originally created in Germany's Fraunhofer Institute.

http://www.hhi.fraunhofer.de/en/depa...sing/overview/

The intention was to create efficient 3D Television signal compression method. It was established that in order to satisfy limited bandwidth compromises such as stereoscopic detail reduction had to be implemented.
The creators also acknowledge that the methods creates 3D artifacts that are not present in original content.

For those who would like to test the MVC codec and it's 3D implementation, here is the source code:

The details loss and artefacts you mention sound exactly like the same compromises that all lossy compressed video formats have and which are not avoidable unless you use raw-source/lossless formats (which are not an option for movie retail and broadcast).

Using the JM encoder to decide on the quality of the format would be one of the worst mistakes to do, the JM AVC reference encoder is known to be old, inefficient and painfully slow, It does not show how good the format can really be (relative to the leading encoders) and I bet the JM MVC one has exactly the same problem, i'll wait for x264 to implement MVC to do the tests with my own content, although I already know what to expect.

From all the comments from specialists I've read, MVC encoding allows theoretically lower bandwidth for the same video quality than dual AVC encoding, or using the same usual logic from all mpeg4 encoders : better quality for the same file size, and considering the 2D + Delta approach it makes perfect sense.
And the video quality of the BluRay3D discs I have in my hands is just impeccable, I have no complaints to make.

Passive 3D, forever !
My Full-HD dual-projector passive polarised 3D setup. (really out of date ! I need to update it some day...)

BlackShark is offline  
post #22 of 74 Old 11-20-2010, 05:22 PM
 
Lee Stewart's Avatar
 
Join Date: Feb 2007
Location: Albuquerque, NM
Posts: 19,369
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 44
Quote:
Originally Posted by icester View Post

Since it's inception in 1971 IMAX totally relied on gimmick
that is an adrenaline WOW effect on viewers due to gigantism of scenes and actors in every movie.
It goes for both in 2D and 3D movies. Additionally IMAX productions of 3D movies are no different than competition and it also uses parallax limit driven stereoscopic content distortion techniques which yield totally unrealistic immersion.

Matt - ALL movies are a gimmick in one form or another.

Quote:


Again MVC 2D + Delta technique was originally created in Germany's Fraunhofer Institute.

http://www.hhi.fraunhofer.de/en/depa...sing/overview/

The intention was to create efficient 3D Television signal compression method. It was established that in order to satisfy limited bandwidth compromises such as stereoscopic detail reduction had to be implemented.
The creators also acknowledge that the methods creates 3D artifacts that are not present in original content.

For those who would like to test the MVC codec and it's 3D implementation, here is the source code:

http://iphome.hhi.de/suehring/tml/download/

Mathew Orman

Matt:

From your link I see this link:

http://www.hhi.fraunhofer.de/en/depa...cast-with-mvc/

And it says nothing about Blu-ray disc. Just broadcast which would mean CBL/OTA/DBS. Substantial difference in the max data rates between those delivery systems and BD.
Lee Stewart is offline  
post #23 of 74 Old 11-21-2010, 03:04 AM
 
icester's Avatar
 
Join Date: Apr 2010
Posts: 1,175
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by Lee Stewart View Post

Matt - ALL movies are a gimmick in one form or another.



Matt:

From your link I see this link:

http://www.hhi.fraunhofer.de/en/depa...cast-with-mvc/

And it says nothing about Blu-ray disc. Just broadcast which would mean CBL/OTA/DBS. Substantial difference in the max data rates between those delivery systems and BD.

Panasonic have adopted MVC 2D + Delta in Blu Ray 3D adding navigation GUIs and exec code capabilities.

Mathew Orman
icester is offline  
post #24 of 74 Old 11-21-2010, 03:15 AM
 
icester's Avatar
 
Join Date: Apr 2010
Posts: 1,175
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by BlackShark View Post

The details loss and artefacts you mention sound exactly like the same compromises that all lossy compressed video formats have and which are not avoidable unless you use raw-source/lossless formats (which are not an option for movie retail and broadcast).

Using the JM encoder to decide on the quality of the format would be one of the worst mistakes to do, the JM AVC reference encoder is known to be old, inefficient and painfully slow, It does not show how good the format can really be (relative to the leading encoders) and I bet the JM MVC one has exactly the same problem, i'll wait for x264 to implement MVC to do the tests with my own content, although I already know what to expect.

From all the comments from specialists I've read, MVC encoding allows theoretically lower bandwidth for the same video quality than dual AVC encoding, or using the same usual logic from all mpeg4 encoders : better quality for the same file size, and considering the 2D + Delta approach it makes perfect sense.
And the video quality of the BluRay3D discs I have in my hands is just impeccable, I have no complaints to make.

Te topic is not about dual AVC but SbS anamorphic vs MVC's 2D + Delta.
Both using the same core compression.

If 2D + Delta File Size = SbS anamorphic File Size then the only difference is nonuniform detail definition in SbS anamorphic and 3D artifacts and reduced low contrast detail in MVC's 2D + delta.

Mathew Orman
icester is offline  
post #25 of 74 Old 11-21-2010, 03:23 AM
 
icester's Avatar
 
Join Date: Apr 2010
Posts: 1,175
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by BlackShark View Post

Using the JM encoder to decide on the quality of the format would be one of the worst mistakes to do, the JM AVC reference encoder is known to be old, inefficient and painfully slow, It does not show how good the format can really be (relative to the leading encoders) and I bet the JM MVC one has exactly the same problem, i'll wait for x264 to implement MVC to do the tests with my own content, although I already know what to expect.

JM MVC is a reference based on free C++ library
and if you want to implement it for real time then you must select the target platform and it's optimal minimum processor power. Most of MVC implementations use assembler and processor based RISC features that are stream compression related.

Mathew Orman
icester is offline  
post #26 of 74 Old 11-21-2010, 03:36 AM
Advanced Member
 
BlackShark's Avatar
 
Join Date: Jun 2010
Location: Montpellier, France
Posts: 580
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 13
Quote:
Originally Posted by icester View Post

If 2D + Delta File Size = SbS anamorphic File Size then the only difference is nonuniform detail definition in SbS anamorphic and 3D artifacts and reduced low contrast detail in MVC's 2D + delta.

And by what kind of miracle would the anamorphic SBS file not be affected by stereo 3D artifacts and reduced lower contrast detail ?
There is still a left and right eye view to encode , there are still 2D artifacts in side by side and if they are different between the views you get stereo 3D artifacts. Also the details are smaller with anamorphic SBS, why would the encoder spare them in the anamorphic SBS and not in 2D+delta.

Passive 3D, forever !
My Full-HD dual-projector passive polarised 3D setup. (really out of date ! I need to update it some day...)

BlackShark is offline  
post #27 of 74 Old 11-21-2010, 03:59 AM
 
icester's Avatar
 
Join Date: Apr 2010
Posts: 1,175
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by BlackShark View Post

And by what kind of miracle would the anamorphic SBS file not be affected by stereo 3D artifacts and reduced lower contrast detail ?
There is still a left and right eye view to encode , there are still 2D artifacts in side by side and if they are different between the views you get stereo 3D artifacts. Also the details are smaller with anamorphic SBS, why would the encoder spare them in the anamorphic SBS and not in 2D+delta.

Because SbS anamorphic is compressed as is, without additional image processing.
2D + Delta requires initial image processing in order to reduce low contrast details. If such is not applied the delta would always be as big as 2D and overall no different in size than original Left + Right.
Additionally when decompressing 2D + Delta the newly formed right image has 3D artifacts in form of 3D surface smooving or flattening.
The only step applied in SbS after decompression is horizontal rescaling using interpolation which makes the nonuniform distribution of detail.

As Fraunhofer pointed out one of the main advantages of 2D + Delta was 2D compatibility. Which in 3D Blu Ray simply means that the Delta is ignored when in 2D Blu Ray playback mode.

Mathew Orman
icester is offline  
post #28 of 74 Old 11-21-2010, 04:37 AM
Advanced Member
 
BlackShark's Avatar
 
Join Date: Jun 2010
Location: Montpellier, France
Posts: 580
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 13
Quote:
Originally Posted by icester View Post

Because SbS anamorphic is compressed as is, without additional image processing.
2D + Delta requires initial image processing in order to reduce low contrast details. If such is not applied the delta would always be as big as 2D and overall no different in size than original Left + Right.
Additionally when decompressing 2D + Delta the newly formed right image has 3D artifacts in form of 3D surface smooving or flattening.

Most of the anamorphic SBS encoding I've performed show exactly the 3D artifacts you are describing. Anamorphic SBS is not immune to it at all, the amount and visibility of these artifacts depend on the bitrate.
The initial processing in order to reduce low contrast detail is part of any lossy video compression scheme, including 2D AVC which you are probably using for Anamorphic SBS.

Saying 2D+delta would be as big as 2D left+right is incorrect. Even if taking uncompressed pictures, the similarities between the left and right eye views mean that on most of the picture, the delta requires less bits to define pixel colour, you can achieve lossless 24bit precision with less than 24 bits of data on average across the pixels of the delta picture. It is one of the advantages of encoding similar pictures through a delta.
Add to that all the motion compensation mechanisms of AVC ported into MVC and the lossy compression (the details reduction which AVC and all others MPEG encoders already use in 2D encodings) which allows the video producer to control the size of the encoded file through the bitrate and you get the significant bitrate reduction requirements claimed by the MVC format.

I maintain that at similar bitrates the less efficient 2D AVC encoding and scaling artifacts present in anamorphic SBS makes the picture look less good than native resolution MVC.

Passive 3D, forever !
My Full-HD dual-projector passive polarised 3D setup. (really out of date ! I need to update it some day...)

BlackShark is offline  
post #29 of 74 Old 11-21-2010, 07:09 AM
 
icester's Avatar
 
Join Date: Apr 2010
Posts: 1,175
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by BlackShark View Post

Most of the anamorphic SBS encoding I've performed show exactly the 3D artifacts you are describing. Anamorphic SBS is not immune to it at all, the amount and visibility of these artifacts depend on the bitrate.
The initial processing in order to reduce low contrast detail is part of any lossy video compression scheme, including 2D AVC which you are probably using for Anamorphic SBS.

Saying 2D+delta would be as big as 2D left+right is incorrect. Even if taking uncompressed pictures, the similarities between the left and right eye views mean that on most of the picture, the delta requires less bits to define pixel colour, you can achieve lossless 24bit precision with less than 24 bits of data on average across the pixels of the delta picture. It is one of the advantages of encoding similar pictures through a delta.
Add to that all the motion compensation mechanisms of AVC ported into MVC and the lossy compression (the details reduction which AVC and all others MPEG encoders already use in 2D encodings) which allows the video producer to control the size of the encoded file through the bitrate and you get the significant bitrate reduction requirements claimed by the MVC format.

I maintain that at similar bitrates the less efficient 2D AVC encoding and scaling artifacts present in anamorphic SBS makes the picture look less good than native resolution MVC.

I have not said anything about compression related artifacts which are common to both layouts and only pointed out the differences due to an extra image processing.

If you do not believe that on every raw pair of stereoscopic images the Delta is as big as left or right then do the subtraction in Photoshop and see the difference.
The black areas would be the common pixels that are identical in left and right image.

Mathew Orman
icester is offline  
post #30 of 74 Old 11-21-2010, 09:41 AM
AVS Addicted Member
 
walford's Avatar
 
Join Date: May 2003
Location: Orange County, CA
Posts: 16,789
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
I understand Matt to be saying that all currently available film based 3D is of poor quality and he refers to this poor qualiy as " a Gimmick".
IMHO the currently used methods to record and play 3D content will continue to exist until some other methods can be developed and be avaiable at prices that are compeitive to "the Gimmick" methods we have today.
walford is offline  
Reply 3D Tech Talk

User Tag List

Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off